Documentation for sdf::Visual::Pose() and sdf::Collision::Pose()

**Original report ([archived issue](https://osrf-migration.github.io/sdformat-gh-pages/#!/osrf/sdformat/issues/185)) by Alejandro Castro (Bitbucket: [amcastro0622](https://bitbucket.org/%7Bf78f4af9-eed7-49c4-b222-5b3dc10d5093%7D/)).**

----------------------------------------
Now the documentation reads:
"Get the pose of the visual object. This is the pose of the visual as specified in SDF"

I believe the pose of a gets specified by default in the parent frame? or is the default the model frame?

This issue is to request to specify in what frame the pose is specified, in the source and [in the online docs](http://sdformat.org/spec?ver=1.6&elem=visual#visual_pose).

Ideally we'd like it to read: "It returns the pose X_LV of the visual frame V in the frame of the parent link L" (or model frame M if that is the case).

Similarly for `sdf::Collision::Pose()`.

Start with the documentation for sdf::Visual::Pose() and sdf::Collision::Pose(), then compare the corresponding visual and collision pose sections in the online SDF specification. Document the reference frame for each returned pose consistently in the source and online documentation, including the parent link or model frame as appropriate.

Written by the indexing model from the issue text.